MODEL PREDIKSI PERUBAHAN PENGGUNAAN LAHAN DENGAN PENDEKATAN SISTEM INFORMASI GEOGRAFIS DAN CELLULAR AUTOMATA MARKOV CHAIN: STUDI KASUS KABUPATEN ROKAN HULU, PROVINSI RIAU

Abstract

Rokan Hulu Regency, Riau Province, is one of the largest oil palm plantation centers in Indonesia. The large growth of palm oil plantations in recent years has threatened the existence of protected areas and increased deforestation rates. This study aims to (1) find out the extent of land use change in Rokan Hulu Regency in 2000-2012, (2) predict land use change in Rokan Hulu regency by 2030. This research uses Geographic Information System approach with prediction model of Cellular Automata (CA) Markov Chain. The driving factors of land use change as input of CA Markov Chain model consist of distance to road, distance to river, distance to disturbance area, distance to Plantation Forest, distance to Forest Clearance Permits and slope data. The results showed that the area of deforestation reached 60.38% from 2000 to 2012, mostly in the northeastern region of Rokan Hulu Regency. The prediction model of land use change has an accuracy of 68.15% based on the kappa index agreement (KIA). By 2030, the area of forest land in Rokan Hulu is predicted to be 3.13%